Tuesday, July 5, 2011

Open Rates

Open rate refers to the number of recipients that opened a specific email and is measured using an embedded HTML code that requests a transparent tracking image from web servers. When a reader opens the email, the ISP used to display the email requests that image, and an ‘open’ is recorded for that particular piece of mail.

The real question is: How much emphasis should be put towards analysis of open rates as a performance metric? Due to problems regarding interpretation in relation to image files, it’s long been debated the accuracy and weight this measure should hold – and some overlook open rates as valuable statistical information.

Conflicts regarding open rates include:

  • Due to the fact that an ‘open’ will only be recorded if the readers ISP is capable of displaying HTML images, if the option is turned off and the recipient chooses to only receive text-only emails there is no way to record the open rates.

  • Although the e-mail has been opened by the recipient there is no guarantee that the receiver read or actively engaged with the email in anyway, all this means is that the tracking image was requested, so you must understand not everyone opening is definitely taking in the information.

  • Some e-mails come equipped with a preview function in which the email is displayed automatically and therefore downloads the tracking image without the receiver ever having viewed the image or even clicked the message.


Open rates do have some positives that should not be overlooked, mainly concerning their use as indicators. How open rates change over time, for example, can tell you a great deal because changes in the open rate reflect real-time changes in your marketing efforts and show problems or successes in regards to how your campaign is running.

When using methods like A/B splits, looking at long term trends, making open rates a base for other metrics, and even comparing campaign results across multiple ISP’s the measure it quite useful.

Open rate can be an effective measure in performance analysis – just be careful how much weight you ultimately give this statistic.

Author: Caitlin Durand
Editor: Yasifur Rahman

No comments: